SGA Data Platform · Ops Monitor

Prototype · sample data 1 critical · 3 warnings

One pane for the whole data platform — governed queries, the PBI/DI bridges, the crosswalk refresh, app traffic, and who touched what. Built to surface problems before anyone reports them. Auto-refreshes every 60s.

Needs your attention derived alerts

Auto-raised from freshness thresholds, canary/outage state, job results, and unusual traffic — so you hear it here first.

Live activity sessions + governed queries

Server-authoritative log: sign-ins and every governed data pull (actor, tool, metric, scope, rows, outcome). This is the audit trail behind the right-to-monitor notice.

TimeUserEventDetailScopeRowsOutcome

Data freshness is anything stale?

Each source vs. its expected cadence. Red = past threshold = numbers on the dashboards are going stale.

SourceLast updatedExpectedStatus

Traffic last 24h

Sign-ins
18 users
Queries
— denied
Avg latency
p95 4.1s
queries / hour · last 24h

Most active 24h

Subsystems

Prototype — representative sample data. In production this is an admin-only tab in sga-one-corp, reading The One's session/access log and the sga-data MCP audit. Wiring + schema in ops-monitor-WIRING.md.
All times US Central · admin-only · every user is notified of the right to monitor at sign-in.